Process for the preparation of the 1'-ethoxycarbonyloxyethyl ester of benzylpenicillin

Abstract

 The invention provides compounds of the general formula

wherein X is -Cl, Br or -OCH₂CH₃, and the preparation thereof. One such compound, 1-bromoethyl ethyl carbonate, is useful as an esterification agent, e.g. for penicillins, and the other such compounds are intermediates in its preparation. The manufacture of formulations containing 1 '-ethoxycarbonyloxy ethyl esters of penicillins is included in the invention.
There are also provided novel compounds of the formula

in which Ph is phenyl and R is CH₃- or C₂H₆-.

Description

[0001] The present. invention concerns certain novel compounds, including 1-bromoethyl ethyl carbonate also known by the alternative name a-bromo diethyl carbonate. This invention also relates to a process for the preparation of 1'-ethoxy carbonyloxy ethyl esters of penicillins, in .which this new potent esterification agent is used. These esters, used as antibiotic pro-drugs, are produced in better yields and higher purity than known before.

[0002] 1-Chloroethyl ethyl carbonate, which is analogous to the new compound of this invention, is a known compound and has been used for esterification purposes in the preparation of drugs of the penicillin series:such as, for example, a-aminopenicillins and penicillin G. Thus, for example, British Patent Specification No. 1,363,506 discloses a process by which a compound of the general formula

wherein R stands for a variety of aliphatic or aromatic radicals, A stands for an amino group, a substituted amino group or a group convertible into an amino group and R" stands for hydrogen or a cation, is reacted inter alia with 1-chloroethyl ethyl carbonate and if required the resulting ester is further reacted to convert group A into a free amino group, to yield finally an ester of the c-amino penicillin series having the formula

[0003] Jt has been found that the resorption of such asters from the intestinal tract is superior to the resorption of the free acids, which means that upon oral administration they yield higher blood concentrations than the corresponding free acid c-aminopenicillin.

[0004] It should be noted that in the context of the said esterification reaction the disclosure in British patent specification No. 1,363,506 refers to a group of compounds

in which X stands for halogen and R¹ stands, inter alia, for the radical

in which radical R⁴ is an alkyl group of 1 to 4 carbon atoms and R is H, CH₃ or C₂H₅. By permutating _R⁴, R and X one can purely on paper arrive, inter alia, at 1 - bromoethyl ethyl carbonate but specifically that compound is not disclosed and was not actually made and is not part of the state of the art.

[0005] To be more specific in the context of said esterification ampicillin, which is D(-) a-amino benzyl penicillin, is an effective broad spectrum antibiotic drug, especially against Gram-positive bacteria. It is widely used all over the world, as a human and veterinary drug. Ampicillin is administered orally, due to its relative stability in the gastric fluids. Absorption of ampicillin from the gastrointestinal tract, however, is far from being complete. Substantial amount of drug does not reach the blood stream. Larger doses are thus required in order to achieve therapeutical levels in the body.

[0006] A solution to this problem is to modify the ampicillin molecule in order to enhance its absorption. One way of doing it is by esterifying the carboxylic group of ampicillin. Absorption of ampicillin esters was found to be much better than the parent compound. Several esters, which are well absorbed, and split off readily the esterifying group in the blood stream, were discovered. As a result, an almost quantitative absorption of ampicillin is achieved.. Some of these ampicillin esters are marketed as human drugs.

[0007] Bacampicillin, which is the 1'-ethoxy carbonyloxy ethyl ester of ampicillin (marketed as the hydrochloride) (formula I):

is being successfully used as an antibacterial drug. It is characterized by an almost quantitative absorption from the digestive tract to the blood stream, and a good tolerance by the gastro-intestinal system. Bacampicillin was first disclosed in British Patent No. 1,363,506.

[0008] According to prior disclosures, bacampicillin was prepared in several ways. One process, disclosed in British patent No. 1,363,506, starts with potassium benzyl penicillinate (formula II), which is treated with 1-chloroethyl ethyl carbonate (CEC, formula III) in aqueous dioxane giving 1'-ethoxy carbonyloxyethyl benzyl penicillinate (formula IV) in less than moderate yields.

[0009] In these formulae X is Cl and Ph is phenyl.

[0010] The British patent mentions the esterification of penicillins with compounds of formula III in which X may be, on paper at any rate, among others, halogen and preferably chlorine or bromine, However, there is no specific disclosure of the use of 1-bromoethyl ethyl carbonate (BEC) nor is there any evidence that this compound has ever been obtained or used.

[0011] An improvement of the esterification reaction is disclosed in British patent No. 1,443,738 which consists in the use of an equivalent amount of a phase transfer catalyst such as tetrabutyl-ammonium hydrogen-sulfate. In the improved process reaction time was shorter and the yield improved. In the second step of the process, the amide group was cleaved, using first phosphorous pentachloride, then an alcohol and finally water. This reaction has to be carried out at sub-ambient temperatures (-30°C or less). The product obtained is 1'-ethoxy carbonyloxy ethyl 6-amino penicillanate (formula V). This compound

was acylated, using either D(-) phenyl glycyl-chloride hydrochloride or sodium (potassium) N-(1-methoxycarbonyl- propen-2-yl)-D(-)-a-amino phenyl acetate, affording bacampicillin. In this process , benzyl penicillin is esterified by CEC and then the synthesis follows the known industrial processes for the manufacturing of ampicillin.

[0012] A second process involves compounds of formula VI, wherein E is an amino group .or a group that can be converted to amino group. -

[0013] Such a compound is first esterified by CEC and then E is converted to amino group yielding bacampicillin. This process, taken as such, claims a very wide range of possible reactants for the preparation of bacampicillin.- A closer look at British patent No. 1,363,506 reveals that, in the preparation of many esters of a-amino penicillins (including bacampicillin), only one such method was employed where E = N3. The a-azido penicillin was esterified with CEC. Hydrogenolysis of the azido penicillin ester gave the corresponding ampicillin ester, e.g. bacampicillin.

[0014] Another procedure U.S. patent 4,072,677, claims the esterification of an N-protected ampicillin by CEC under phase transfer catalysis conditions, followed by removal of the protecting group. In fact, U.S. Patent Specification No. 4,072,677 discloses a group of compounds R₂X in which X may stand, inter alia, for a halogen atom such as bromine or iodine and R₂ may, inter alia, be the ethoxycarbonyloxyethyl (i.e. 1-ethyl ethyl carbonate) radical so that here too by permutation one may arrive purely on paper, inter alia to 1-bromoethyl ethyl carbonate, but again that specific compound has not been disclosed and was not actually made and is not part of the state of the art.

[0015] In all the above mentioned processes, the esterification of the penicillin molecule was carried out by CEC. As mentioned, this is a known compound and was first reported in the chemical literature in 1889. CEC is commercially available. From the family of 1-haloethyl ethyl carbonates (formula III), this is the only known compound.

[0016] 1-Chloroethyl ethyl carbonate was first prepared by Muller by chlorination of ethylchloroformate in direct sunlight and reacting the resulting α-chloroethyl chloroformate with ethanol:

see H. Müller, Ann. 252, 50 (1889).

[0017] Attempts to produce the corresponding bromo compound failed. Thus Landenburg and Wichelhaus, Ann. 152, 163 (1869) reported an attempt to brominate directly diethyl carbonate. However the reaction resulted in degradation of diethyl carbonate to ethyl bromide, bromal and carbon dioxide. Nesmeyanow et al. reported in Izvest. Akad. Nauk SSSR, Otdel, Khim. Nauk 631 (1949) the action of bromine on alkyl chlorocarbonate. The reaction resulted in degradation, yielding, inter alia, 1,2-dibromoethane.

[0018] It has long been known that the C-Br bond is more labile than the C-Cl bond and that in consequence 1-chloroethyl ethyl carbonate is less reactive in many reactions than the corresponding bromo compound, 1-bromoethyl ethyl carbonate. Thus the rate of esterification with 1-chloroethyl ethyl carbonate is slow and as many as three equivalents of the chloro compound are required to complete the esterification. Moreover, drastic conditions such as long reaction time, high temperature and the use of catalysts such as phase transfer catalysts are required. Where such drastic conditions are applied to esterifications of sensitive compounds such as, for example, penicillins or compounds of the a-aminopenicillin series, they may cause partial degradation of the sensitive β-lactam system. This indeed explains the known fact that when 1-chloroethyl ethyl carbonate is used to esterify ampicillin the overall yields are yields are relatively low. Indeed, in general reduced yields and impure products are obtained. In all the processes for preparing bacampicillin, the yields are relatively low and many procedures give material of inferior quality.

[0019] It is also known that many chloro compounds are lachrymators and skin irritants. Any esters produced with CEC should be subjected to extensive and thorough purification operations to remove traces of CEC.

[0020] Because of the relative lability of a C-Br bond as compared to a C-Cl bond, it was to be expected that an esterification reaction would proceed much more smoothly with 1-bromoethyl ethyl carbonate (BEC) than with 1-chloroethyl ethyl carbonate (CEC), requiring less reagent, shorter reaction time, lower reaction temperature and no catalyst while at the same time producing better yields. Where the acid subjected to esterification is sensitive such as, for example, an acid of the a-aminopenicillin series, shortening of the reaction time and lowering the reaction temperature have favourable effects on the yield of the desired final product in that the degradation of the sensitive S-lactam moiety is minimized.

[0021] Thus, in view of all the various deficiencies of 1-chloroethyl ethyl carbonate, there has been a longfelt want to replace it by 1-bromoethyl ethyl carbonate (BEC), having the formula VII:

[0022] However, BEC has never been reported in the chemical literature. Furthermore, published reports taught that bromination of diethylcarbonate or ethyl chloroformate, the two possible precursors for BEC, gave only degradation products. No BEC was obtained in these reactions.

[0023] In accordance with the present invention this longfelt want for BEC has for the first time been fulfilled.

[0024] Accordingly, the invention provides a compound of the general formula

wherein X is Cl, Br or OCH₂CH₃.

[0025] The direct bromination of diethyl carbonate, ethyl chloroformate or ethyl bromoformate runs against the difficulty that the a-position which has to be brominated is greatly deactivated by the adjacent electron withdrawing O-CO-O grouping and this is presumably the reason for which hitherto-all attempts to produce 1-bromoethyl ethyl carbonate have failed.

[0026] In accordance with the present invention it has been found that _BEC is readily obtainable in good yields by either of two methods:

a) a free radical type reaction performed with a member of the group of diethyl carbonate, ethyl chloroformate and ethyl bromoformate-in-which the brominating agent must be present in an amount not exeeding the stoichiometric quantity where ethyl chloroformate or ethyl bromoformate is used as starting material. The products 1-bromoethylchloroformate and 1-bromoethylbromoformate, which are novel compounds, may be reacted with ethanol to yield BEC ; and

b) a substitution type reaction in which the chlorine . in 1-chloroethyl ethyl carbonate is replaced by bromine with an excess of bromide salt..

[0027] In the free radical type reactions the free radical initiation may be effected by light e.g. UV and visible light, if desired with moderate heating, or by means of radical initiator compounds, e.g. organic peroxides, such as benzoyl peroxide, azobisisobutyronitrile or any other known free radical initiator. The free radical type bromination reaction may be effected with elementary bromine or with a brominating agent such as, for example, l,3-dibromo-5,5-dimethyl oxazolidine-2,4-dione.

[0028] The chlorine-bromine substitution reaction is performed in an organic aprotic polar solvent such as, for example, acetone.

[0029] The various manners by which _BEC can be produced in accordance with the present invention are depicted by the following reaction schemes:

[0030] In these schemes, (i) to (iv) represent the free radical method while (v) represents the chlorine-bromine substitution method.

[0031] 1-Bromoethylchloroformate and 1-bromoethyl- bromoformate, obtainable in method (iv) above, are also novel compounds forming part of the present invention This invention includes, in general, a process for the preparation of 1-bromoethylchloroformate in which the starting material is ethylchloroformate, and a process for the preparation of 1-bromoethylbromoformate in which the starting material is ethylbromoformate.

[0032] The foregoing different methods for the preparation of BEC in accordance with the invention will now be discussed briefly.

(i) Free radical initiation with light (photobromination) using elementary bromine

[0033] The exact wavelength of the light used for the irradiation of the reaction mixture is not critical. Several sources emitting UV and visible light have been used such as a regular tungsten lamp, a quartz-iodine lamp and a high pressure mercury lamp, and all of these caused bromine to react'with diethyl carbonate in the desired manner. A careful examination of the strong lines of a high pressure mercury lamp from 313 nm to 578 nm showed that all ofthem produced good results. What determines the efficiency of the irradiation is the rate of absorption of light by bromine and not the wavelength of the irradiation. The most efficient light absorption was found in the region of 390-410 nm but bromination could also be carried out with other wavelengths, the only requirement being that the light should be absorbed by bromine.

[0034] The bromination of diethyl carbonate in the presence of light can be performed at various temperatures. Reactions were carried out successfully with good yields and product quality in the range of 60-126°C at the beginning of the reaction and at the end of the reaction the temperature rose as high as 140°C. Diethyl carbonate boils at 126°C and hence forms an upper limit for the reaction temperature unless high pressure equipment is used. Below 60°C the reaction becomes extremely slow and the time factor makes low temperature reactions impractical. The preferred temperature range, considering .time and light efficiency, is from 80-126°C.

[0035] The bromination reaction may be carried out in a non-hydroxylic solvent inert to bromine and hydrogen bromide such as carbon tetrachloride, dibromoethane, tetrachloro- ethene, chlorobenzene and the like. A preferred solvent is 1,1,2-trichloro trifluoroethane (Freon 113, Trademark).

[0036] When a solvent is used, the bromine is preferably added gradually to a boiling solution of diethyl carbonate in the solvent. Usually the heat produced by the lamp is enough-to cause the reaction mixture to boil. The ratio of solvent can vary within a broad range, e.g. in the case of carbontetrachloride from 0.4 - 4.2, the preferred ratio being about 1.5.

[0037] Preferably at no time should a large amount of free bromine be present in the reaction mixture. Therefore the bromine is preferably added at such a rate that the solution will be only lightly coloured by the added bromine. Where the concentration of bromine is increased, this results in a degradation of both reactant and reaction product and the yield of BEC is considerably reduced. The actual rate of bromine addition so that the above colour proviso be maintained will depend in each case on the parameters of the system, such as lamp intensity, wavelength available, mixing efficiency, _'etc. and must be determined empirically by the operator in each case.

[0038] A total low amount of bromine results in a low conversion of diethyl carbonate with little side reactions. An excess of bromine gives rise to the formation of by-products such as bis-(1-bromoethyl)carbonate, i.e. a dibromination product. Compromising between high conversions of starting material and low yields of polybrominated products, it has been found in accordance with the present invention that in the free radical method not more than the theoretical (stoichiometrical) amount of bromine should be added and .preferably not more than 85% of that amount. At such amounts polybrominated products account for less than 10% of the reaction mixture. Where only 40% of the theoretical amount of bromine is added polybromination is negligible.

[0039] It is also possible to conduct the above reaction in the absence of any solvent. Thus it is possible to reflux diethyl carbonate and irradiate it, and to add bromine slowly .so that the colour of the mixture remains yellow to light orange. When 40% of the theoretical amount.of bromine is added the reaction mixture consists of almost only diethyl carbonate and 1-bromoethyl carbonate while the amount oi polybrominated by-products is negligible. Also in this case it is possible to add more bromine and to achieve a higher conversion of diethyl carbonate but then by-products are beginning to form.'

[0040] The free radical bromination of diethyl carbonate in the absence of a solvent may also be carried out below the boiling point, e.g. at 90°C, in which case the reaction will proceed much slower but the yield and quality of the products are not affected.

[0041] In all the foregoing cases of light induced free radical bromination of diethyl carbonate, heating and irradiation are continued until all bromine is consumed. The mixture is then subjected to fractional distillation under reduced pressure which produces two to three major fractions, namely solvent if any, unreacted diethyl carbonate and finally c-bromodiethyl carbonate.

(ii) Free radical initiation with initiator compounds usino elementary bromine

[0042] Any organic compound that is capable of free radical formation upon heating can be employed, examples being organic peroxides such as benzoyl peroxide and azobisisobutyronitrile

[0043] The reaction with a free radical initiator compound is carried out in the same range of temperatures, concentrations of reactants and solvents, if any, as described under (i) above. Temperatures of 85°C were found to be useful when the reactions are carried out in the dark with azobisisobutyronitrile as initiator. Unlike photobromination as described under (i) above the addition of bromine need in this case not be very slow. In fact one can add up to 30% of the theoretical amount of bromine at the beginning together with the free radical iniator and then heat the reaction mixture. The balance of the bromine is then added gradually when the reaction is sufficiently in progress. If sufficient amount of the initiator is present bromine is consumed and a good conversion of diethyl carbonate to BEC takes place. However, yields of BEC by this method are not as good as by the photochemical method described under (i) above.

(iii) Free radical bromination with brominating aqents

[0044] The bromination with elementary bromine as described under (i) and (ii) above produces hydrogen bromide which apart from being an ecological nuisance also has a strong destructive effect on organic carbonates. It may therefore be preferable to use organic brominating agents, e.g. such that contain a bromine atom linked to a nitrogen atom. Such bromine is readily liberated with free radical initiation and becomes available for the bromination of another molecule. Examples of such brominating agents are N-bromosuccinimide, N-bromocaprolactam, N-bromoacetamide, 1,3-dibromo-5,5-dimethylimidazolidine-2,4-dione.

[0045] . Such brominating.agents can replace elementary bromine both in the light initiated.free radical bromination described under (i) above and the free radical bromination using free radical initiator compounds described under (ii) above. When proceeding by this method one may operate with or without a solvent. In any event it is preferred to keep a threefold excess of diethyl carbonate over the brominating agent, or in other words to use not more than about 33-35% of the theoretical amount of the brominating agent, in order to avoid the liberation of free bromine.

(iv) Free radical bromination of ethylchloroformate or ethylbromoformate

[0046] As mentioned hereinbefore this method was tried .by Nesmcyanow et al and failed. In spite of this discouraping report in the literature it was found in accordance with the present invention thai the bromination proceeds smoothly if the specific conditions stipulated in accordance with the present invention are maintained. The bromination may be carried out by any of the methods described under (i) to (iii) above. The compound 1-bromoethylchloroformate obtained as intermediary in this process, is a novel compound.

(v) Production from 1-chloroethyl ethyl carbonate by chlorine-bromine substitution

[0047] 1-Chloroethyl ethyl carbonate is commercially available. The substitution reaction proceeds in an aprotic polar solvent and the bromide source is a bromide salt, e.g. lithium bromide or an organic salt such as tetraalkylammonium bromide, 'e.g. tetrabutylammonium bromide.

[0048] As indicated earlier, BEC is a very potent esterification agent for the preparation of 1'-ethoxycarbonyloxy ethyl esters. It is especially useful when the esterification is carried out on salts of relatively unstable and sensitive acids. This is the case with most β-lactam antibiotics. The esters of benzyl penicillin and N-protected a-amino penicillins (ampicillin, amoxycillin) may be prepared in excellent yields and high purity under mild conditions (low temperatures and short reaction time).

[0049] For example, according to British-patent No. 1,363,506 1'-ethoxycarbonyloxyethyl benzyl penicillinate was prepared by reacting potassium benzyl penicillinate with CEC in aqueous dioxane at room temperature for 66 hours. The ester was obtained at 33% yield with an unspecified degree of purity. An improvement was reported later in British patent No. 1,443,738 according to which the rather expensive tetrabutyl ammonium benzyl penicillinate was prepared and this salt was then subjected to the esterification with CEC in boiling acetone for 6 hours. The ester was obtained in an overall yield of 65% and in 90% purity.

[0050] An unexpected significant improvement has nowbeen achieved in accordance with'the present invention by using BEC instead of CEC for the esterificaticn. Thus, potassium benzyl penicillinate was reacted with BEC in moist acetone at 40°C for 5 hours to give l'-ethoxycarbonyloxy ethyl benzyl penicillinate in a quantitative yield and 99% purity. This is a dramatic increase in yield and product quality using at the same time milder reaction conditions as shown in accordance with the present invention. The same improvements were observed in accordance with the invention in the esterification of other penicillins. Levels of toxic contaminants are reduced.

[0051] The product obtained from the esterification of potassium benzyl penicillinate with BEC gives a starting material for many 1'-ethoxycarbonyloxy ethyl esters of known semi-synthetic penicillins. Thus, the ester having formula IV was cleaved by a known sequence of chemical reactions to afford the ester of 6-amino penicillinanic acid having formula.V. Acylation of V with the appropriate agent affords the penicillin ester. Thus treating V with D(-) phenyl glycyl chloride hydrochloride gives bacampicillin (formula I), in good yields. The high purity of IV obtained from BEC gave bacampicillin in higher yields and better purity than reported before, when CEC was used.

[0052] The esterification of potassium benzyl penicillinate was carried out in non-hydroxylic polar solvents. Examples are acetone, 2-butanone, dimethylsulfoxide and acetonitrile. The preferred solvent is acetone. The reaction is best carried out in the presence of an acid absorber such as sodium bicarbonate The temperature range for the reaction is between 20° to 50° but the range of 35 - 45° is preferred. Reaction is best carried out in moist solvent (ca. 4% water) in order to accelerate the esterification (enhancing the solubility of the potassium salt). The reaction rate can also be accelerated by the addition of a catalytic amount of a quaternary ammonium salt such as tetrabutyl ammonium bromide, but the addition of such a salt is not an essential part of the invention.

[0053] A new process for the production of bacampicillin, using BEC as esterification agent, is part of the present invention.

[0054] Direct esterification of ampicillin by BEC gives bacampicillin in lower yield. The product is contaminated with several compounds of which one of the major contaminants is believed to be of the formula VIII

[0055] The presence of an amino group, another possible site for reaction, has'to be avoided. As a result, the esterification has to be carried out on a derivative of ampicillin with a protected amino group. BEC will then react only with the carboxylic acid. After the esterification is complete, removal of the protecting'group affords bacampicillin. The amino protecting group must be removed under very mild conditions, due to the sensitivity of the penicillin nucleus. Protecting groups usually employed in peptide chemistry are not suitable. They do not fulfill the condition of an easy removal. We found that protection in the form of enamine by reaction of ampicillin with alkyl acetoacetate is suitable. The protecting group is removed under mild acidic conditions. Any such-compound of formula IX wherein M is sodium, potassium or tri (lower) alkyl ammonium cation and wherein R₁ is a methyl or ethyl group, will be referred to hereinafter as "ampicillin Dane salt":

[0056] 

A solution of IX (M = HN(C₂H₅)₃; R1 = CH₃) is obtained easily by reacting ampicillin trihydrate with triethylamine and methylacetoacetate. The nature of the solvent is not so important. Acetone, dichloromethane, ethyl acetate, isopropanol, acetonitrile, and methyl acetoacetate are mere examples for suitable solvents. The reaction is carried out in the temperature range of 10° - 50° preferably at 20-40°. Usually an excess of alkyl acetoacetate is used in order to achieve a corplete protection of the amino group.

[0057] A crystalline potassium salt of IX (R₁⁼ CH₃; M = K) can be prepared by treating a solution of IX with a potassium salt soluble is the organic solvent used.

[0058] The other crystalline form of ampicillin, anhydrous ampicillin, reacts in the same way but much longer reaction time is needed duc to lower dissolution rate.

[0059] Based on the foregoing observations and discoveries, the invention provides a process for the preparation of a 1'-ethoxycarbonyloxyethyl ester of a penicillin having the formula:

in which A is a phenyl group, phenoxy group or 4-hydroxyphenyl group and B is hydrogen, an amino group or a protected amino group, wherein a compound of the formula

wherein A and B are as in formula X and Z is hydrogen or a cation selected from the group of alkali metal, tri (lower alkyl) ammonium and tetra (lower alkyl) ammonium, is reacted with 1-bromoethyl ethyl carbonate in an organic solvent and-when B is a protected amino group the protecting group is split off to yield a primary amino group.

[0060] If desired, the reaction is carried out in the presence of an acid acceptor.

[0061] Also if desired the resulting ester may be subjected to treatment for splitting off the acyl residue

to yield 2-ethoxycarbonyloxyethyl ester of G-amino penicillanic acid and the latter is then acylated with a different acyl group, as is known per se. In this way one 1'-ethoxycarbonyloxy ethyl ester of a semi-synthetic penicillin according to the invention may be converted into another by replacing the acyl residue.

[0062] By way of example, the reaction of penicillin G (formula II or the corresponding sodium salt) with BEC in acetone at 40° proceeds smoothly to give the ester having formula IV in quantitative yield and 99% pure. If desired, catalytic amounts of a tertiary amino or quaternary ammonium salt may be added, e.g. triethyl amine or- tetrabutyl ammonium bromide. In this way the reaction may be accelerated.

[0063] In a similar manner salts of penicillin V. (formula XI; A = phenoxy, B = hydrogen) react to give the corresponding ester (formula X: A = phenoxy, B = hydrogen).

[0064] These esters are a suitable source for the preparation of bacampicillin. Cleavage of the acyl group by known procedures (e.g. phosphorous pentachloride, alcohol and water) affords the ester of the formula V. This compound is acylated by an appropriate derivative of phenyl glycine to give bacampicillin.

[0065] Other suitable candidates for the preparation of bacampicillin are compounds having the formula IX. These compounds are esterified smoothly by BEC to give .compounds having formula XII_I

in which Ph is phenyl and R is methyl or ethyl. This compound named bacampicillin Dane "salt" (BDS) yields bacampicillin upon treatment with dilute aqueous acid, e.g. hydrochloric acid.

[0066] ADS of formula XIII are novel compounds. When these compounds are obtained as intermediates in the preparation of bacampicillin they may or may not be isolated.

[0067] In a similar way the I'-ethoxycarbonyloxy ethyl ester of amoxycillin can be prepared.

[0068] Preferably an excess of 30-120% of BEC is used for the esterification in order to complete the reaction. Usually, an excess of 50-100% of BEC is used. Under these conditions, esterification is complete within several hours. Excess of BEC is destroyed by water at room temperature. The N-protecting group is then removed by a careful addition of aqueous hydrochloric acid. Gare must be taken during the acid addition lest the pH will drop too low. Making the solution too acidic results in partial decomposition of the penicillin. A final, stable pH of ca. 2.2 was found most satisfactory for complete removal of the _N-protecting group while keeping the penicillin molecule intact.

[0069] . The esterification can be carried out equally well on a solid ampicillin. Dane salt or a solution of an ampicillin Dane salt. In the latter case the product obtained in the protection step is directly reacted with BEC. Usually ADS is also not isolated and a "one pot" process from ampicillin to bacampicillin is carried out.

[0070] The process according to the invention has several advantages over the known processes for the preparation of 1-ethoxycarbonyloxy esters of penicillins. Taking, for example, the case of bacampicillin, the new process gives bacampicillin in higher yield than disclosed before and of very good quality. The chemical operations needed to be executed in the process are simple. For instance there is no need to work at sub-ambient temperatures (-30°C and below). The mild conditions of the process lead to minimal degradation of the sensitive β-lactam compounds. No toxic materials are involved in the process. In the known processes, conversion of IV to V necessitates the use of N,N-dimethylaniline or quinoline. These are toxic materials and they should be removed. Their removal from the final product is very difficult. The present invention completely avoids the use of toxic compounds.

[0071] In general the invention provides a novel esterification agent which may be used, inter alia, for the preparation of bacampicillin and similar drugs. This enables improvements over existing processes for the preparation of bacampicillin. Using a new esterification agent, a marked improvement over the present know-how is established.

[0072] Thus, the invention includes the use of 1-bromoethyl ethyl carbonate as an esterification agent.

[0073] The invention also includes a method for the manufacture of a pharmaceutical or veterinary formulation, which method comprises formulating a pencillin ester which has been prepared by the process of the invention for pharmaceutical or veterinary use, respectively, optionally including a diluent, carrier or excipient in the formulation and/or optionally rendering the formulation in unit dosage form. The skilled man may apply his knowledge and experience in preparing such formulations.

[0074] Of course, the present invention is not limited to the use of BEC in esterification only. The invention provides BEC for the first time'and embraces all uses thereof.

[0075] The invention is further illustrated by the following Examples to which it is not limited. The properties of the product 1-bromoethyl ethyl carbonate were determined on the product of Example 1. In Examples 2-7, 11 and 12 it was established that the product was identical with that of Example 1. Thereafter, Examples 13 to 21 show how 1-bromoethyl ethyl carbonate may be used in the synthesis of semi-synthetic penicillins.

Example 1

[0076] A mixture of 354 g (3 mole) diethylcarbonate and 700 ml carbon tetrachloride was stirred and irradiated externally with a 1.5 kw iodine - quartz lamp. The heat from the lamp brought the mixture to reflux. Bromine (412.8 g, 2.58 mole) was added slowly through a dip-pipe below the surface of the mixture over a period of 15-17 hours.

[0077] The colour of the solution was kept light red throughout the reaction. The temperature in the solution was 84-85°C. Hydrogen bromide evolved copiously during the addition.

[0078] The reaction mixture was cooled, washed with aqueous sodium bisulfite and dried over magnesium sulfate. Carbon tetrachloride was distilled at atmospheric pressure.

[0079] The rest was fractionally distilled at 30 mm Hg pressure The first fraction consisted of unreacted diethylcarbonate, the second fraction boiled at 90-95° (30 mm pressure) was 1-bromoethyl ethyl carbonate. The residue in the pot was mainly bis-(1-bromoethyl) carbonate. The yield was about 52% calculated on bromine.

[0080] The following physical data of 1-bromoethyl ethyl carbonate:

GC analysis showed it to be a pure compound.

Example 2

[0081] 848.2 g (7.18 mole) of diethyl carbonate and 420 ml 1,1,2-trichloro-trifluoro ethane ("Freon" 113 Trade Mark) were stirred and irradiated by a 150 w high pressure mercury immersion lamp. The mixture was heated by the lamp and soon began to boil at-75°. Bromine (936 g, 5.85 mole)' was added slowly through a dip-pipe at such a rate to keep the solution lightly coloured. Addition took 4 1/4 hours. During the addition the temperature in the flask rose steadily from 75° to 94°. HBr evolved during the addition. The reaction was cooled to 40° and dry nitrogen was passed until fumes of HBr did no longer evolve. The reaction was subjected to fractional distillation as described above. The yield of isolated 1-bromoethyl ethyl carbonate was 62%.

Example 3

[0082] 585 g Diethylcarbonate (4.92 mole) was heated by an external 1500 w iodine quartz lamp. The mixture was allowed to reflux (temperature 125°) by the heat evolved from the lamp. 318.2 g (2.0 mole) of bromine is added slowly during 3 hours through a dip-pipe at such a rate that the solution colour was yellow to light orange. The temperature dropped to ca. 120° at the beginning of the addition, but rose again slowly. When all the bromine was consumed the temperature was 137°. The reaction mixture was cooled and treated as described in Example 2. Fractional distillation afforded 1-bromoethyl ethyl carbonate in a 68.5% yield. The earlier fraction contained about 6.5% more of 1-bromoethyl ethyl carbonate.

Example 4

[0083] 6 g (51.0 mmole) diethylcarbonate, 100 ml carbon tetrachloride and 28 g (17.4 m.mole) of bromine were refluxed in the dark. 1 g of azobisisobutyronitrile was added and the mixture was heated for 2 hours. Analysis of the product showed that 28% of diethylcarbonate were converted to 1-bromoethyl ethyl carbonate.

Example 5

[0084] The experiment described in Example 4 was repeated but using 0.1 g of azobisisobutyronitrile. The yield of 1-bromoethyl ethyl carbonate was 6% after 2 hours. Addition of a second portion of 0.1 g azobisisobutyronitrile an heating for another hour increased the yield to 9%.

Example 6

[0085] 29.25 g diethylcarbonate and 5 g 1,3-dibromo-5,5- dimethylimidazolidine-2,4-dione were heated to 95°. 0.5 g azobisisobutyronitrile were added. The solid disappeared immediately. The reaction was heated to 110° for 1 hour. The mixture was cooled and filtered.

[0086] The filtrate was treated by aqueous sodium bisulfite and dried. GC analysis showed ca. 5% conversion of diethyl carbonate to 1-bromoethyl ethyl carbonate. Yield based on 1,5-dibromo-5,5-dimethyl imidazolidin-2,4-dione was 36%.

Example 7

[0087] A mixture of 195 g diethylcarbonate and 500 ml of 1,1,2-trichlorotrifluoroethane was irradiated externally by a 1.5 kw iodine-quartz lamp. The mixture soon refluxed by the heat evolved from the lamp. The temperature in the flask was 60°C. 79.0 g of 1,3-dibromo-5,5-dimethylimidazoline-2,4-dione was added in small.portions during 6 hours. The mixture was irradiated and refluxed for another 20 minutes and cooled. Solids were filtered. The solution was fractionally distilled first at atmospheric pressure and then under. vacuum. 1-Bromoethyl ethyl carbonate was distilled at 91° at 20 mm Hg pressure. The fraction weighed 60 g and was shown to be 98% pure, by GC.

Example 8

[0088] Ethyl chloroformate (271.6 g, 2.5 mole) was irradiated with an external 1.5 kw iodine-quartz lamp i and purged with N₂ throughout the reaction. Upon boiling, 1.0 mole of bromine was gradually added over 10 hours, so that the colour of the solution was kept between yellow and light orange. The mixture was then fractionally distilled. 1-Bromoethyl chloroformate was isolated in the fraction that boiled at 50° under 20 mm Hg pressure in 60% yield. Its formula was confirmed by NMR (doublet of 3H at 2.1 ppm and quartet of 1H at 6.5 ppm). GC analysis showed it to be 96% pure. 3t is a colourless liquid with pungent odour that fumes in air. It decomposes readily on exposure to humid air.

Example 9

[0089] A mixture of 32.5 g ethylchloroformate and 0.5 g azobisisobutyronitrile was refluxed in the dark. 40.8 g of bromine were added slowly during 1 hour. After 2 hours from the beginning of the addition the reaction mixture was analyzed. A complex mixture resulted which contained 11% of ethylchloroformate and 47% of 1-bromoethyl chloroformate.

Example 10

[0090] A mixture of 10.8 g ethylchloroformate, 50 ml carbon tetrachloride, 18 g of N-bromosuccinimide and 0.5 g azobisisobutyronitrile was refluxed for 30 minutes. The product was analyzed and was found to contain 3% 1-bromoethylchloroformate.

Example 11

[0091] A mixture of 37.4 g 1-bromoethylchloroformate, 20 ml dichloromethane and 26 g anhydrous sodium carbonate were stirred at room temperature. 12 ml of absolute ethanol were added dropwise during 45 minutes. 3 portions of 4 ml absolute ethanol were added after 2, 3 and 4 hours. After 5 hours the mixture was filtered. Dichloromethane and excess ethanol were evaporated at room temperature. The residue was distilled in vacuo go give 23 g of 1-bromoethyl ethyl carbonate.

Example 12

[0092] A mixture of 15.25 g 1-chloroethyl ethyl carbonate (0.1 mole), 34.8 g lithium bromide (0.4 mole) and 200 ml acetone was refluxed for 2 hours. Acetone was evaporated and the residue was stirred with 200 ml dichloromethane. The solids were filtered and dichloromethane evaporated. The residual liquid was analyzed to contain 65% of 1-bromoethyl ethyl carbonate and 35% 1-chloro ethyl ethyl carbonate.

Example 13

[0093] A mixture of potassium benzyl penicillinate (20 g), sodium bicarbonate (18.15 g), water (2.4 ml) and acetone (60 ml) was heated to 40°C. 1-Bromoethyl ethyl carbonate (21.2 g) was added during 30 minutes. The reaction mixture was stirred for 4.5 hours at 40°. Water (60 ml) was added and the reaction mixture stirred for 1 hour at ca. 30°C. Ethyl acetate (60 ml) was added and the mixture was filtered. The organic phase was separated and washed with a saturated sodium bicarbonate solution and 10% sodium chloride solution. The mixture was dried and the solvent evaporated under reduced pressure. 1'-Ethoxycarbonyloxyethyl benzyl penicillinate was obtained in quantitative yield and 99% purity (assayed by HPLC and mercurometric titration).

Example 14

[0094] A mixture of potassium benzyl penicillin (20 g), 1-bromoethyl ethyl carbonate (19.15 g), sodium bicarbonate (18.15 g), tetrabutyl ammonium bromide (1 g) and acetone (60 ml) was heated at 45°C for 2.5 hours. Water (60 ml) was added and the reaction mixture stirred for 1 hour and then ethyl acetate (60 ml) was added. The organic phase was separated, washed with a sodium bicarbonate solution (2 x 60 mi) and a 20% sodium chloride solution.

[0095] The organic phase was dried over sodium sulphate and decolourized with charcoal. The solvent was evaporated under reduced pressure to.yield 22.1 g of 1-ethoxy carbonyloxyethyl benzyl . penicillinate 99% pure by G.C. and mercurometry (91.1% total yield).

Example 15

[0096] A mixture of potassium benzyl pencillinate (20 g), 1-bromoethyl ethyl carbonate (19.15 g), sodium bicarbonate (18.15 g), tetrabutyl ammonium bromide (1 g) and acetone (60 ml) was stirred at 45°C for 2.5 hours. The mixture was treated in the same way as described in Example 1 to give l'-ethoxycarbonyloxy ethyl benzyl penicillinate 97% pure, in quantitative yield.

Example 16

[0097] 1'-Ethoxycarbonyloxy ethyl benzyl penicillinate (100 g), obtained as described in the preceding examples, was dissolved in methylene chloride (375 ml). N,N-Dimethyl aniline (61 ml) was added and the mixture was cooled to -80°C. Phosphorous pentachloride was added in three portions of 31 g taking care that the temperature does not rise above -60°C.

[0098] The mixture was stirred at -70° for

hours. Cold methanol (140 ml) was added slowly keeping the temperature below -55°C. Stirring was continued at -70°C for 2 hours. Water (500 ml) and petroleum ether (450 ml) were added and the phases separated. The organic layer was washed with water (500 ml). The combined aqueous fractions were combined and basified to pH 7 with ammonia, keeping the temperature at 10°C or below.

[0099] The organic phase was separated and triturated with petroleum ether (3 x 100 ml) and the solvent decanted each time. The residual thick oil was dissolved in methylene chloride (240 ml). Sodium bicarbonate (30 g) and water (3 ml) were added. The mixture was cooled to -5°C. D(-) phenylglycyl chloride hydrochloride (37.2 g) was added and the mixture was stirred for 2 hours at 0°C. The mixture was filtered from solids, methylene chloride was evaporated. Butyl acetate (240 ml) and isopropanol (30 ml) were added. Bacampicillin crystallized slowly at 0°C. The solid was filtered, washed and dried. The yield of .bacampicillin was 68.2 g.

Example 17

[0100] To 1_-bromoethyl ethyl carbonate (145.8 g) in acetone (720 ml) at 35°C was added sodium bicarbonate (160 g) and Ampicillin Dane potassium salt (240 g). The suspension was heated at 40°C for 5 hours. Water (720 ml) was then added and stirring continued for an additional hour, when ethyl acetate (720 ml) was added. The organic layer was separated and washed with 20% aqueous sodium chloride (720 ml), then concentrated under reduced pressure (below 35°C) to a thick residue.

[0101] To the residue was added acetone (720 ml), water (24 ml) and then dropwise a solution of conc. HCl (36 ml) diluted with water (12 ml). An additional amount of HCl (1:1, 7.5 ml) was then added until a constant pH of about 2.2 was obtained. Magnesium Sulphate (40 g) was then added,'stirred for 10 minutes and filtered. To the solution was added butyl acetate (400 ml), evaporated in vacuo to remove acetone and butyl acetate (about 320 ml). The residue was diluted with butyl acetate (880 ml) and part of the solvent evaporated under reduced pressure.

[0102] The solid was filtered off, washed with butyl acetate (120 ml) and ethyl acetate (120 ml), dried at 40° in vacuo, giving pure bacampicillin hydrochloride.(176.2g, a yield of 71.3%). The product conformed with U.S.A. Pharmacopoeia requirements and exhibited an assay of 93.7%.

Example 18

[0103] A mixture of ampicillin trihydrate (15 g), methyl- aceto acetate (5.65 g), triethylamine (4.05 g) and acetone (7.5 ml) was stirred at 40°C for 3 hours. Acetone (30 ml) and sodium bicarbonate (12.4 g) were added. 1-Bromoethyl ethyl carbonate (14.6 g) was added dropwise.during 1 hour. Stirring was continued at 40°C for 1 hour. Water (37.5 ml) was added and the reaction was stirred at 20°C for 1 hour. Ethyl acetate (37.5 ml) was added and the phases separated.

[0104] The organic phase was washed with 20% sodium chloride ; solution: Acetone (7.5 ml) was added and a solution of ca. 16% hydrochloric acid was added slowly until a stable pH of 2.2 was obtained. The addition takes

- 2 hours. The solution was dried over anhydrous sodium sulfate and filtered. Butyl acetate (45 ml) was added and the solution was evaporated in vacuo to remove most of the solvents (bath temperature 35°C). More butyl acetate (50 ml) was added and partially distilled. The resulting slurry was cooled for several hours and filtered to give 11.6 g bacampicillin.

Example 19

[0105] A mixture of ampicillin Dane potassium salt (45.6 g), 1-bromoethyl ethyl carbonate (36.0 g), sodium bicarbonate (38.4 g) and acetonitrile (260 ml) is stirred for 3 hours at 40°C. The mixture is filtered and the cake is washed with acetonitrile (2 x 20 ml). Water (20 ml) is added and the mixture is stirred for 1 hour. Most of the solvent is removed by evaporation at 35 - 40°C, under reduced pressure. Acetone (70 ml) and water (4.5 ml) are added. A mixture of concentrated hydrochloric acid (6.85 ml) and water (2.3 ml) is added slowly in order to keep the pH at 2 - 2.5. If, at the end, the pH is not stable, a small additional amount of 1:1 hydrochloric acid is added until a stable pH in the range 2-2.5 is obtained. The solution-is dried and filtered. Butyl acetate (180 ml) and the solution is distilled under reduced pressure to a volume of ca. 40 ml. More butyl acetate (270 ml) is added and the product is allowed to crystallize at ca. 5°C. The mixture is filtered to give 32.5 g bacampicillin.

Example 20

[0106] Ampicillin Dane potassium salt (30.0 g), sodium bicarbonate (20.0 g), 1-bromoethyl ethyl carbonate (16.9 g) and acetone (90 ml) were stirred at 40° for 5 hours.

[0107] The reaction mixture was cooled to room temperature and ethyl acetate (90 ml) and 20% sodium chloride solution (100 ml) were added. The mixture was filtered and the phases separated. The organic layer was dried and concentrated. Final concentration under high vacuum gave the 1'-ethoxycarbonyloxy ethyl ester of ampicillin Dane salt (ADS). An analytical sample was crystallized from ether. The compound.was characterized by its NMR and IR spectra.

Example 21

[0108] A reaction mixture of acetone (24.2 ml), methyl acetoacetate (8.4 ml) and triethylamine (9 ml) was heated to 40°C. Ampicillin trihydrate (24.2 g) was added and the mixture maintained at 40°C for 1.5 hours.

[0109] To the reaction mixture was then added 1-bromoethyl ethyl carbonate (17.75 g), sodium bicarbonate (20 g) and acetone (48.4 ml), and heated at 40°C for 5 hours. Thereafter deionized water (72.4 ml) was added to decompose excess of BDEC and stirred until a TLC test showed the absence of BDEC.

[0110] Ethyl acetate (72.4 ml) was then added, the mixture .filtered, the organic layer wahsed with 20% sodium chloride (72 ml). To it was added dropwise a solution of.HCl 1:1 at room temperature to obtain a final stable pH of 2.2, which was maintained for about 1.5 hours.

[0111] To the solution was added butyl acetate (75 ml) concentrated to dryness at reduced pressure at about 40°C. An additional quantity of butyl acetate (75 ml) was then added and 5-10 ml distilled out under the same conditions.

[0112] The solid was filtered, washed with butyl acetate (10 ml) and ethyl acetate (20 ml), and dried at 40°C in a vacuum oven.

[0113] Yield of Bacampicillin HC1 19.75 (65%9 Assay 95.6% and passes all requirements of the U.S. Pharmacopeia.

Claims

1. A compound of the general formula

wherein X is Cl, Br or OCH₂CH₃.

2. A process for the production of a compound as claimed in claim 1 comprising either: (a) brominating a compound of formula
CH₃ - CH - 0 - CO - X,
wherein X has the same meaning as in claim 1, under conditions conducive of free radical formation, the amount of bromine made available in the bromination being not more than the stoichiometric quantity; or (b) reacting a compound as claimed in claim 1 in which X is Cl or Br with ethanol to produce a compound as claimed in claim 1 in which X is OCH₂CH₃; or (c) reacting 1-chloroethyl ethyl carbonate with an excess of a bromide salt in an organic polar solvent.

3. A process as claimed in claim 2, wherein the bromination in (a) is carried out under irradiation.

4. A process as claimed in claim 2 or claim 3, wherein heating is effected during the bromination in (a).

5. A process as claimed in claim 2,wherein (a) is carried out in the heat in the presence of at least one organic compound capable of forming free radicals in the heat.

6. A process as claimed in claim 5, wherein the organic compound is a peroxide or azobisisobutyronitrile.

7. A process as claimed in any one of claims 2 to 6, wherein elementary bromine or an organic brominating agent is used for the bromination in (a).

8. A process as claimed in claim 7, whereih the brominating agent is N-bromosuccinimide, N-bromocaprolactam, N-bromoacetamide or 1,3-dibromo-5,5-dimethyl- imidazolidine-2,4-dione.

9. A process as claimed in any one of claims 2 to 8, wherein the starting material in (a) is dissolved in a non-hydroxylic inert' solvent, e.g. carbon tetrachloride, dibromoethane, chlorobenzene, tetrachloroethane or 1,1,2-trichlorotrifluorethane.

10. A process as claimed in any one of claims 2 to 8, wherein the starting material in (a) is brominated without the use of a solvent.

11. A process as claimed in claim 2, wherein in (c) the solvent is acetone.

12. A process as claimed in claim 2 or claim 11, wherein the bromide salt used in (c) is lithium bromide or a quaternary ammonium bromide, e.g. tetrabutyl ammonium bromide.

13. The use of 1-bromoethyl ethyl carbonate as an esterification agent.

14. A process for the preparation of penicillin 1'- ethoxycarbonyloxyethyl esters which comprises reacting an optionally protected penicillin with 1-bromoethyl ethyl carbonate.

15. A process as claimed in claim 14 and for the preparation of a penicillin ester having the formula:

in which A is a phenyl group, phenoxy group or 4-hydroxyphenyl group and B is hydrogen, an amino group or a protected amino group, wherein a compound of the formula

wherein Z is hydrogen or alkali metal, tri (lower alkyl) ammonium or tetra (lower alkyl) ammonium, is reacted with 1-bromoethyl ethyl carbobate in an organic solvent and when B is a protect-d amino group the protecting group is split off to yield a primary amino group.

16. A process as claimed in claim 15 carried out in the presence of an acid acceptor.

17. A process as claimed in claim 15 or claim 16, wherein A is phenoxy, or A is phenyl and B is hydrogen.

18. A process as claimed in claim 15 or claim 1₆, wherein a compound of the formula:

in which R₁ is methyl or ethyl and Ph is phenyl, is reacted with 1-bromoethyl ethyl carbonate and the resulting ester is treated with a dilute aqueous acid to split off the CH₃-C=CHCOOR₁ group so as to produce bacampicillin.

19. A compound of the formula:

in which Ph is phenyl and R is CH₃ - or C₂H₅ -.

20. A method for the manufacture of a pharmaceutical or veterinary formulation, which method comprises formulating a penicillin ester which has been prepared by a process as claimed in any one of claims 14 to 18 for pharmaceutical or veterinary use, respectively, optionally including a diluent, carrier or excipient in the formulation and/or optionally rendering the formulation in unit dosage form.